Laminar Distribution of Neurochemically-Identified

Neurosci. Bull. www.neurosci.cn https://doi.org/10.1007/s12264-018-0275-x www.springer.com/12264

ORIGINAL ARTICLE

Laminar Distribution of Neurochemically-Identiﬁed Interneurons and Cellular Co-expression of Molecular Markers in Epileptic Human Cortex

2 1 1 3 1 Qiyu Zhu • Wei Ke • Quansheng He • Xiongfei Wang • Rui Zheng • 3 3 4 4 Tianfu Li • Guoming Luan • Yue-Sheng Long • Wei-Ping Liao • Yousheng Shu1

Received: 23 January 2018 / Accepted: 20 May 2018 Ó The Author(s) 2018

Abstract Inhibitory GABAergic interneurons are funda- layers IV and VI. Neurons with these markers constituted mental elements of cortical circuits and play critical roles *7.2% (PV), 2.6% (SST), 0.5% (TH), 0.5% (NPY), and in shaping network activity. Dysfunction of interneurons 4.4% (CCK) of the gray-matter neuron population. Double- can lead to various brain disorders, including epilepsy, and triple-labeling revealed that NPY neurons were also schizophrenia, and anxiety. Based on the electrophysio- SST-immunoreactive (97.7%), and TH neurons were more logical properties, cell morphology, and molecular identity, likely to express SST (34.2%) than PV (14.6%). A interneurons could be classiﬁed into various subgroups. In subpopulation of CCK neurons (28.0%) also expressed this study, we investigated the density and laminar PV, but none contained SST. Together, these results distribution of different interneuron types and the co- revealed the density and distribution patterns of different expression of molecular markers in epileptic human cortex. interneuron populations and the overlap between molecular We found that parvalbumin (PV) and somatostatin (SST) markers in epileptic human cortex. neurons were distributed in all cortical layers except layer I, while tyrosine hydroxylase (TH) and neuropeptide Y Keywords Interneuron Á Epilepsy Á Human cortex Á Cell (NPY) were abundant in the deep layers and white matter. type Á Immunostaining Á Parvalbumin Á Somatostatin Á Cholecystokinin (CCK) neurons showed a high density in Tyrosine hydroxylase Á Neuropeptide Y Á Cholecystokinin

Electronic supplementary material The online version of this Introduction article (https://doi.org/10.1007/s12264-018-0275-x) contains supplementary material, which is available to authorized users. In the cerebral cortex, non-pyramidal GABAergic interneu- & Yousheng Shu rons are involved in cortical information-processing and [email protected] high-order cognitive functions. Though non-pyramidal cells (20%–30%) are much less numerous than the main 1 State Key Laboratory of Cognitive Neuroscience and Learning and IDG/McGovern Institute for Brain Research, output neurons, pyramidal cells (70%–80%) [1–3], in the Beijing Normal University, Beijing 100875, China whole population of cortical neurons, they are more 2 College of Pharmaceutical Sciences, Brain Institute, Capital extensive and complex. Interneurons can be classified Medical University, Beijing 100069, China depending on their morphology, intrinsic membrane prop- 3 Department of Neurosurgery, Epilepsy Center, Sanbo Brain erties, and synaptic connectivity and dynamics. Distinct Hospital of Capital Medical University, Beijing Key Labo- interneuron subtypes can be also identified by the expres- ratory of Epilepsy, Epilepsy Institution, Beijing Institute for sion of specific molecular markers, such as parvalbumin Brain Disorders, Beijing 100093, China (PV), somatostatin (SST), tyrosine hydroxylase (TH), 4 Institute of Neuroscience and Department of Neurology of vasoactive intestinal polypeptide, ionotropic serotonin the Second Affiliated Hospital of Guangzhou Medical 5-hydroxytryptamine 3a receptor, nitric oxide synthase, University, Key Laboratory of Neurogenetics and Chan- nelopathies of Guangdong Province and the Ministry of cholecystokinin (CCK), and neuropeptide Y (NPY) [4–6]. Education of China, Guangzhou 501260, China Although some of them (e.g. CCK) are also expressed by a 123 Neurosci. Bull. subpopulation of glutamatergic pyramidal cells [7], neu- density is high in layers II, III, and VI, and in the white rons labeled with these markers are most likely to be matter of human cortex [36]. TH is a molecular marker of GABAergic cells in the neocortex [5, 8, 9] and possess midbrain dopaminergic neurons and is the rate-limiting distinct electrophysiological and morphological features. enzyme of dopamine synthesis. Some cortical cells also For example, PV-expressing neurons show a fast-spiking express this enzyme and may reflect a unique cell type in firing pattern and send axons to innervate the perisomatic the neocortex [37]. TH neurons in the human cortex are regions of pyramidal cells, while SST-containing neurons mainly located in deep layers and are fusiform, bipolar, or show a low-threshold spiking firing pattern and innervate multipolar [38, 39]. Early immunostaining experiments the distal apical dendrites of pyramidal cells [10–13]. revealed the expression of CCK in a subpopulation of Because of the fundamental role of GABAergic interneu- GABAergic neurons in the neocortex [40]. Some of the ron in providing inhibitory control of cortical network, CCK-expressing neurons in the hippocampus are basket changes in interneuron circuitry and alterations of cells targeting the perisomatic regions of pyramidal cells GABAergic transmission in the cortex can lead to disorders [41, 42]. Selective loss of hippocampal CCK-containing of cognition and emotion, such as schizophrenia, anxiety, boutons and thus a decrease in GABA release may cause and epilepsy [14–17]. epilepsy in an animal model of temporal lobe epilepsy An epileptic seizure is a paroxysmal alteration of [40, 43]. CCK neurons in the human cortex are also function caused by excessive, hyper-synchronous discharge positive for calretinin or reelin [7]. Since a particular cell of neurons and abnormal network activity in the brain. type may express a combination of markers, it is of interest Although numerous pathogenic conditions can result in to determine whether PV and SST cells also express NPY, epilepsy along with brain dysfunction [14, 18, 19], its TH, and CCK, and determine whether cells positive for one pathophysiology is generally considered to be a distortion of these markers are a subpopulation of PV or SST cells. of the normal, well-balanced excitation (E) and inhibition In this immunohistochemical study, we performed (I) in the brain [20]. A genetic or acquired E–I imbalance double- and triple-labeling to analyze the distribution can result from changes at many levels, from genes and patterns of the above molecular markers of interneurons subcellular signaling cascades to neural circuits. GABAer- and their co-localization in cortex from patients. gic interneurons are critical circuit elements in the cortex, providing inhibition in cortical networks, and thus con- tribute significantly to the E–I balance. Alterations in their Materials and Methods distribution and density in the cortex, as well as changes in the co-localization of different molecular markers in We used epileptic tissues removed from 9 patients during interneuron subtypes may reflect the mechanisms underly- brain surgery. All were associated with secondary epilepsy ing brain diseases. Previous studies have revealed an caused by various pathological conditions. The cortical association between hippocampal GABAergic interneurons areas removed were mainly from the frontal or temporal and the generation of epilepsy [21]. Changes in GABA lobe (Table 1). Non-epileptic peri-tumor tissues were production or GABA receptor expression have been found obtained from 3 patients (frontal, temporal, or parietal in epileptic tissues [22, 23]. However, the distribution and lobe). The Ethics Committee of Beijing Sanbo Hospital, co-localization patterns of different molecular markers for Capital Medical University, China approved all studies. GABAergic interneurons in the human epileptic cortex We have complied with all relevant ethical regulations need to be further explored. relating to the use of resected human brain tissue in Among cortical interneurons, PV- and SST-expressing research. The clinical investigations were conducted cells are the most abundant cell types [5, 24]. In the human according to the Declaration of Helsinki. Informed consent cortex, PV neurons including chandelier cells and large was given by all participants or their parents or legal basket cells [25] comprise * 20% of all GABAergic guardians. neurons [26]; SST neurons are distributed unevenly across Cortical tissues were removed during the course of the human cortex [25, 27, 28]. PV and SST neurons play neurosurgery for the treatment of patients with important roles in the generation of cortical network intractable epilepsy. Before the surgery, epileptogenic activity, such as gamma and beta oscillations [29–31], as regions were identified by video-EEG recording and the well as seizure-like activity [32]. NPY is a neuropeptide removed regions were associated with significant abnormal produced by certain types of neurons throughout the brain spiking. Since we sought to explore the laminar distribution and by secretory cells of other systems [33, 34]. In the patterns of interneuron subtypes, we chose parts of the neocortex, NPY is expressed in a subpopulation of removed tissue blocks showing complete cortical layers GABAergic neurons and is involved in brain disorders (from layer I to white matter). All tissues were immersed in including seizure activities [33, 35]. The NPY neuron 3% paraformaldehyde (PFA) and 3% sucrose in 0.1 mol/L 123 Q. Zhu et al.: Interneuron Distribution in Human Cortex

Table 1 Patient information No. Age, Duration of Etiology Seizure Previous Drug treatment Surgical Surgical resection sex epilepsy frequency treatment results area

116m,7 m FCD 2a Several VNS Oxcarbazepine, Sodium val- Seizure- Right temporal F times/d proate, Clonazepam free lobe 219m,12 m Sturge-Weber 2–4/m None Oxcarbazepine, Topiramate, Seizure- Left anterior M syndrome Clonazepam free temporal lobe 314y,14 y Trauma 3/m None Sodium valproate Seizure Left frontal lobe M decrease 416y,6 y FCD 1b 10/m IEI Carbamazepine, Sodium Seizure- Left anterior F valproate free temporal lobe 539y,30 y FCD 2b 4-5/w None Sodium valproate No record Right frontal lobe M 628y,9 y FCD 1b 2/m IEI Oxcarbazepine, Sodium No record Left anterior M valproate temporal lobe 728y,20 y FCD 1b Several None Carbamazepine, Sodium Seizure- Left temporal M times/d valproate free lobe 88y,8 y FCD 1b 10/w IEI Oxcarbazepine, Clonazepam Seizure Left anterior M decrease temporal lobe 98y,0.5 y FCD 1b 3–4/w None None No record Left anterior M temporal lobe 10 62 y, – Meningioma – – – – Right frontal lobe F 11 47 y, – Melanoma – – – – Left parietal lobe M 12 62 y, – Breast cancer – – – – Left temporal F lobe m, month; y, year; d, day; w, week; F, female; M, male; focal cortical dysplasia; FCD 1b, defective horizontal lamination with lamina-speciﬁc neuronal paucity [44]; FCD 2b, dysplastic megalocytic neurons, balloon cells, delaminated cortical architecture and abnormal glial cells, mixed with neurons and glia that appear histologically normal [44]; IEI, intracranial electrode implantation; Sturge-Weber syndrome, neurocutaneous disorder with angiomas that involve the leptomeninges; VNS, vagus nerve stimulation.

phosphate buffer (PB, 80 mmol/L Na2HPO4, 16 mmol/L complete cortical layers (from pia to white matter) and at NaH2PO4, pH 7.4) at 4 °C for 2 h. They were then stored in least 1000 NeuN-stained cells. For the co-localization of 30% sucrose in PB overnight at 4 °C for cryoprotection. cell markers (double- and triple-staining), we selected Tissues were cut into 30-lm-thick sections on a freezing regions of similar size but without NeuN staining. Images microtome at - 20 °C. Sucrose and residual PFA were were captured on a laser scanning confocal microscope washed out in 0.01 mol/L PBS buffer (in mmol/L: 8 (A1?, Nikon, Japan) with 109 and 209 objectives. The

Na2HPO4, 1.6 NaH2PO4, and 145 NaCl), the sections were acquisition parameters were carefully adjusted to linearly then pre-incubated in 0.5% Triton X-100 in PB for 30 min, display the fluorescence signals and ensure that they fell in and blocked in 5% bovine serum albumin and 0.5% Triton the maximum dynamic range of the detectors. To generate X-100 in PB for 1 h at room temperature (22 °C). Z-stacks for each marker and each cortical field, we Sections were then incubated overnight at 4 °C and two acquired 3 images (3.5-lm intervals) at a magnification of more hours at room temperature in 0.1% Triton X-100 109 and 6 images (1.5-lm intervals) at 209. Original containing primary antibodies (Supplementary Table S1). images were processed with ImageJ (National Institutes of After washing with 0.01 mol/L PBS, sections were Health, USA) to adjust the brightness and contrast, and incubated for 2 h at room temperature in 0.1% Triton then exported as TIFF images. We used MatLab (2014a, X-100 containing secondary antibodies. Mathworks, USA) and ImageJ to calculate the total number We chose to image non-successive sections (* 2mm9 of cells in regions of interest. To count the number of cells 2 mm) for each experiment so that no cell on the surface labeled by a particular marker, we set a threshold of would be over-counted. We collected images from 3 fluorescence intensity and only those above the threshold regions of interest in separate cortical sections from each were counted. In our experiments, it was clear that the patient. The regions of interest for cell counting contained labeled cells showed a much higher fluorescence intensity

123 Neurosci. Bull. than the background. Positive cells were also determined Results by soma size. For instance, NeuN-labeled cells with a soma size between 5 and 35 lm in diameter were counted. Specificity of Antibodies We identified cortical layers using the following charac- teristics of different layers, assisted by a MatLab code that We performed control experiments to confirm the speci- determined the soma size and density of NeuN-labeled cells. ficity of the primary and secondary antibodies (Supple- Positively-labeled cells in layer I were sparse. Compared mentary Table S1). We followed the protocols described in with layer III, layer II had a higher cell density but the cells the Methods but with some changes in the procedures. To were smaller. Layer IV had a higher cell density than its assess the specificity of the secondary antibodies, for each neighboring layers (III and V). Layer V had the largest primary antibody (NeuN, PV, SST, CCK, NPY, and TH), pyramidal cells. Layer VI cells were smaller than layer V we applied all three secondary antibodies. We found that cells. The border between layer VI and the white matter was only the corresponding secondary antibody showed posi- determined by a sharp decrease in cell density. tive signals (Fig. 1A, SST primary antibody). Next, we

Fig. 1 Control experiments assessing antibody specificity. A Goat Note that most of the labeled cells were positive for both primary anti-SST was used as the only primary antibody and followed by the antibodies. C Double-labeling of mouse and rabbit anti-TH. All three secondary antibodies. Only one channel showed positive signals labeled cells were positive for both primary antibodies. Cortical (Alexa 647 anti-goat). B Double-labeling of mouse and goat anti-PV. tissues were from patient #9 (8 years old). 123 Q. Zhu et al.: Interneuron Distribution in Human Cortex replaced the primary antibody with bovine serum albumin across cortical layers (Fig. 2 and Tables 2, 3, 4 and 5). We and applied all the three secondary antibodies and observed found that PV-ir neurons comprised 7.16% ± 0.54% no staining. In addition, we performed double-labeling (mean ± SEM) of NeuN-labeled gray-matter neurons using two distinct primary antibodies for PV (goat and (n = 26,759 cells from 6 patients) and located in all mouse anti-PV) or TH (rabbit and mouse anti-TH) and cortical layers except layer I, with layer IV displaying the assessed the overlap of labeled cells. The percentages of highest density (Fig. 2A–D and Tables 2, 3). And the cells positive for both antibodies were 78% in goat and density of PV-ir neurons in gray matter was 34 ± 7 per 90% in mouse anti-PV-labeled cells (Fig. 1B). For TH mm2 (n = 1632 PV cells). Moreover, PV-ir neurons were staining, the two primary antibodies labeled the same rarely found in the white matter. The PV-ir neurons were population of cells (completely overlapped, Fig. 1C). most likely basket cells or chandelier cells, as suggested in previous studies by their somatic morphology and axonal Laminar Distribution and Density of PV and SST arborization [48]. PV-ir fibers were densely distributed Neurons across all cortical layers including layer I. We also performed similar labeling in non-epileptic tissues (from In immunohistochemical experiments, we used NeuN as patients #10–12) and found the proportion of PV-ir cells the neuronal marker since it is expressed by most neurons among all cortical neurons (6.99% ± 0.25%, n = 10,258) throughout the central nervous system and has been used as and their density in gray matter (34 ± 5 per mm2, n = 766) a neuronal marker in previous studies [45–47]. Based on were similar to the values obtained from epileptic tissues, double-staining for NeuN and different neurochemical indicating no substantial change of PV-ir cells under markers, we determined the total number of neurons in the epileptic conditions. regions of interest as well as the distribution and density of SST-ir neurons constituted 2.60% ± 0.31% of NeuN- interneuron subpopulations (Fig. 2A). In these experi- labeled gray-matter neurons (n = 35,761) and were dis- ments, we used tissues from patients #1, 3, 4, 7, 8 and 9 tributed mainly in layers IV, V, VI and the border between (Table 1). We performed staining using antibodies to all gray and white matter (Fig. 2E–G and Tables 2, 3). The the molecular markers in cortical sections from each density of SST-ir neurons in gray matter was 16 ± 5 per patient. We also analyzed the cell density (number of cells mm2 (n = 793). SST-ir neurites were mainly located in per mm2) in different layers and the white matter. deep layers; straight and long fibers crossing layers and Labeling with antibodies revealed the distribution perpendicular to the white matter were often observed patterns of neurons immunoreactive (ir) for PV or SST (Fig. 2E and F).

Fig. 2 Laminar distribution of PV-ir and SST-ir neurons. A NeuN (black, patients 1, 3, 4, 7, 8, and 9) and their density in each layer immunostaining showing the identification of cortical layers I to VI (blue, patients 7–9). E Distribution of SST-ir neurons in the right and the white matter. Tissue was from the left frontal lobe of patient temporal lobe of patient #1 (1.3 years old). F Higher magnification of #3 (14 years old). B PV labeling in the same section as in A. C Higher the boxed area in E. G As in D, but for SST-ir cells. Error bars denote magnification of the boxed area in B. D Group data showing the SEM. percentage of PV-ir cells among the total PV-positive population 123 Neurosci. Bull.

Table 2 Percentages of positive cells in each layer I II III IV V VI wm

PV ep 1 0 14.08% 23.94% 34.32% 13.38% 7.75% 3.52% 3 0 11.40% 26.47% 34.93% 19.49% 7.72% 0 4 0 16.67% 28.04% 41.27% 9.26% 3.97% 0.79% 7 0 20.28% 20.56% 39.43% 9.86% 9.30% 0.56% 8 0 12.96% 16.05% 33.95% 23.46% 9.26% 4.32% 9 0 18.80% 22.07% 38.15% 11.17% 8.45% 1.36% Mean 0 15.70% 22.86% 37.01% 14.44% 7.74% 1.76% SEM 0 3.17% 3.93% 2.78% 5.27% 1.80% 1.60% PV non-ep 10 0 15.81% 19.76% 36.36% 19.37% 7.11% 1.58% 11 0 14.05% 27.57% 43.24% 9.73% 4.86% 0.54% 12 0 16.00% 23.00% 39.00% 12.00% 7.00% 3.00% Mean 0 15.29% 23.44% 39.53% 13.70% 6.32% 1.71% SEM 0 0.88% 3.20% 2.83% 4.12% 1.04% 1.01% SST ep 1 0 6.82% 12.50% 11.36% 26.17% 26.14% 17.05% 3 0 7.53% 9.68% 23.30% 21.15% 22.22% 16.13% 4 0 19.21% 19.49% 20.90% 16.10% 14.41% 9.89% 7 0 8.28% 10.19% 22.29% 22.93% 20.38% 15.92% 8 0 6.45% 8.60% 24.73% 21.51% 18.28% 20.43% 9 0 7.81% 10.94% 23.44% 23.44% 20.83% 13.54% Mean 0 9.35% 11.90% 21.00% 21.88% 20.38% 15.49% SEM 0 4.45% 3.60% 4.47% 3.06% 3.59% 3.23% TH ep 1 0 3.70% 7.41% 14.81% 25.93% 29.63% 18.52% 3 0 0 7.14% 7.21% 35.66% 35.71% 14.29% 4 0 0 5.26% 15.79% 21.05% 31.58% 26.32% 7 0 0 5.26% 15.79% 31.58% 26.32% 18.42% 8 0 0 8.33% 8.33% 33.33% 33.33% 16.67% 9 0 0 10.42% 22.92% 35.42% 20.83% 10.42% Mean 0 0.62% 7.30% 14.14% 30.50% 29.57% 17.44% SEM 0 1.39% 1.79% 5.24% 5.32% 4.88% 4.84% NPY ep 1 0 0 0 12.33% 14.33% 20.00% 53.33% 3 0 0 0 15.79% 21.05% 21.05% 42.11% 4 0 0 0 15.38% 7.69% 19.23% 57.69% 7 0 0 0 16.00% 16.00% 20.00% 48.00% 8 0 0 0 7.15% 14.28% 35.71% 42.86% 9 0 0 0 12.50% 16.67% 33.33% 45.83% Mean 0 0 0 13.19% 15.00% 24.89% 48.30% SEM 0 0 0 3.09% 3.97% 6.87% 5.59% CCK ep 1 0 15.07% 13.70% 34.25% 8.22% 24.66% 4.11% 3 0 6.45% 13.98% 32.26% 16.13% 27.96% 3.23% 4 0 21.43% 24.29% 30.00% 7.14% 15.71% 1.43% 7 0 11.49% 12.16% 32.43% 12.16% 30.41% 1.35% 8 0 11.35% 20.87% 30.81% 14.05% 28.65% 0 9 0 9.87% 11.16% 32.62% 12.02% 32.19% 2.15% Mean 0 12.61% 16.03% 32.06% 11.62% 26.60% 2.05% SEM 0 4.69% 4.83% 1.36% 3.12% 5.39% 1.34% ep: epileptic; non-ep: non-epileptic; wm: white matter.

123 Q. Zhu et al.: Interneuron Distribution in Human Cortex

Table 3 Cell density in NeuN-labeled cells (counts per mm2) Table 3 continued I II III IV V VI wm gm I II III IV V VI wm gm

PV-ep 9 0 19 19 38 18 54 2 29 1 0 33 25 47 19 10 2 23 Mean 0 19 19 38 19 41 1 23 3 0 31 52 66 39 15 0 34 SEM 0 8 10 7 7 15 1 7 4 0 32 30 63 20 6 1 29 ep: epileptic; non-ep: non-epileptic; gm: gray matter; wm: white 7 0 50 51 70 28 25 1 44 matter. 8 0 52 40 74 51 18 4 43 9 0 39 36 54 23 22 3 33 Laminar Distribution and Density of NPY, TH, Mean 0 40 39 62 30 16 2 34 and CCK Neurons SEM 0 9 10 9 12 7 1 7 PV non-ep Next, we investigated the distribution patterns of TH-ir, 10033378849322 37 NPY-ir, and CCK-ir neurons (Fig. 3 and Tables 2, 3). TH-ir 11049457620131 39 (Fig. 3A, B, C, F) and NPY-ir (Fig. 3D, G) neurons were 12 0 33 41 66 16 9 1 27 distributed mainly in layers V and VI and the white matter, Mean 0 38 41 77 28 18 1 34 taking up 0.53% ± 0.14% (V and VI; n = 35,448) and SEM 0 8 3 9 15 10 1 5 0.51% ± 0.06% (white matter; n = 26,859) of the total SST ep cortical neurons. The density of TH-ir and NPY-ir neurons 1 0 5 5 6 13 13 4 8 in the gray matter were 3 ± 1(n = 144) and 2 ± 1 per mm2 3 0 7 6 14 15 21 6 13 (n = 132), respectively. Most of the TH-ir neurons were 4 0 20 26 34 25 25 9 23 bipolar or bitufted cells (Fig. 3A, C), and many NPY-ir 7 0 14 15 24 24 25 15 19 neurons emitted two main dendrites in different directions 8 0 13 7 14 18 16 7 13 (Fig. 3D). The TH and NPY antibodies also labeled a mass 9 0 10 15 23 34 29 9 18 of long fibers in layer I and the white matter. In layer I, these Mean 0 12 12 19 22 22 8 16 fibers were horizontal and parallel to the pia. TH-ir fibers SEM0579763 5were present in the superficial layers (Fig. 2B). It should be TH ep noted that most of these TH fibers may come from 1 0113693 4subcortical dopaminergic or noradrenergic neurons [39]. 3 0022582 3 CCK-ir neurons were concentrated in layers IV and VI, 4 0012562 3and a much lower density was observed in other layers. 7 0024884 4They constituted 4.38% ± 0.47% of all cortical neurons 8 0011331 1(n = 28,563), and their density in the gray matter was 2 9 0023875 323 ± 7 per mm (n = 1434). Because the neurites of CCK- Mean 0 1 2 3 6 7 3 3 ir neurons were only weakly labeled, their distribution SEM0111221 1pattern was unclear. The majority of CCK-ir neurons NPY ep possessed a round or oval soma (Fig. 3E). Previous 1 0002247 1experiments have revealed that a subpopulation of pyra- 3 0001464 2midal cells express CCK [7]. We also found a similar 4 0003375 2pattern (except for samples from the infant patient #1); 7 000669132however, the fluorescence immunosignals were relatively weaker than in non-pyramidal cells. Since we focused on 8 00025147 4 cortical interneurons, we only included the strongly-labeled 9 0002428 2 non-pyramidal neurons for data analysis. Mean 0 0 0 3 4 7 7 2 SEM0002143 1 Co-localization of NPY, CCK, and TH with PV CCK ep and SST 1 0 15 11 33 7 23 1 16 3 0 8143314282 21 Although many molecular markers have been used to 4 0 18 11 36 26 26 1 14 identity different interneuron types, none can actually 7 0 19 16 33 21 58 1 24 define a cell type. Given that two or more molecular 8 0 36 41 52 28 57 0 34 markers can be expressed in a single cell type, we next

123 Neurosci. Bull.

Table 4 Percentage of positive 1 (%) 3 (%) 4 (%) 7 (%) 8 (%) 9 (%) Mean (%) SEM (%) cells among NeuN-labeled cells in gray matter of epileptic tissue PV 6.55 7.84 6.99 7.94 6.71 6.92 7.16 0.54 SST 2.16 2.65 3.12 2.44 2.43 2.81 2.60 0.31 TH 0.23 0.58 0.53 0.65 0.54 0.67 0.53 0.15 NPY 0.40 0.59 0.53 0.49 0.56 0.51 0.51 0.06 CCK 4.48 5.26 3.76 4.37 4.42 4.01 4.38 0.47

Table 5 Percentages of cells co-expressing PV or SST among the We also examined the co-expression among NPY, TH, total TH-positive cell population in different patients and CCK (Fig. 7) in double-labeling experiments. Patient number PV?TH?/TH? SST?TH?/TH? Although both TH-ir and NPY-ir neurons were located in deep layers and the white matter, we found no overlap 1 26.1% (6/23) 13.0% (3/23) between these cells (Fig. 7A). No co-localization between 3 9.67% (3/31) 38.7% (12/31) NPY and CCK (Fig. 7B) or TH and CCK (Fig. 7C) was 4 12.5% (2/16) 43.8% (7/16) found. These results suggest that these chemical markers 5 8.33% (1/12) 40.0% (14/35) identify distinct non-overlapping cell types in the neocortex. Tissues from patients #2, 4, and 6 were used in these sought to clarify the overlap between PV-ir and SST-ir experiments. cells and those containing NPY-ir, CCK-ir, or TH-ir. Several distinct subpopulations of NPY-containing neurons have been reported in rat and monkey cortex, one of Discussion which co-expresses SST [49–51]. Consistent with these reports, our results revealed that 97.7% ± 1.5% of NPY-ir Although it is very important to understand the neural interneurons (n = 43) in epileptic human cortex also substrates of human cognitive functions, few studies have expressed SST, and these co-expressing neurons were examined the cell types and their distribution in cortical 38.2% ± 3.1% of all SST-ir neurons (n = 110). Since NPY circuits. Previous work in adult human cortical tissues have neurons were only present in layers IV, V, VI, and the revealed the density and distribution patterns of certain white matter, the co-expressing neurons were mainly types of neurons [25, 39, 52]. In this study, we investigated located in the deep layers and absent from the superﬁcial the cell density and laminar distribution of neurons layers (Fig. 4A–C). In sharp contrast, we found no co- expressing PV, SST, CCK, NPY, and TH in epileptic localization between PV-ir and NPY-ir (Fig. 4D). In this human cortex. Our results indicated that PV, SST, and particular experiment, we examined the co-expression in CCK cells were located in cortical layers II to VI, while tissues from patients #1–6. TH-ir and NPY-ir cells were preferentially located in deep Our experiments showed that 34.2% ± 7.3% of TH-ir layers from IV to the white matter. In addition, we found neurons (n = 105) were SST- positive and 14.6% ± 5.6% subpopulations of neurons labeled by two markers: TH-ir neurons (n = 82) were PV-positive (Table 6). In this PV?TH, PV?CCK, SST?NPY, and SST?TH. In con- part of work, we used the tissues from patients #1, 3, 4, and trast, we found no overlap between PV and NPY, SST and 5 (1, 3, and 4 for triple-staining, and 5 for double-staining). CCK, NPY and TH, NPY and CCK, and TH and CCK As shown in Table 5, the percentages of TH-positive cells (Tables 5 and 6). co-expressing PV or SST varied between patients, possibly We used epileptic tissue to investigate the cell density resulting from age differences. The staining shown in and distribution of interneurons types. It remains to be Fig. 4 was performed in tissues from patients #3 (Fig. 5A) further determined whether speciﬁc cell types are vulner- and #1 (Fig. 5B). able to epileptic conditions. Early examination has We found that 28.0% ± 5.4% of CCK-ir neurons revealed a loss of PV-positive cells in human cortical (n = 321) co-expressed PV in double-staining experiments tissue [22] and the hippocampal CA1 region with temporal- (Fig. 6A). Interestingly, although CCK-positive cells were lobe epilepsy [53]. However, it has also been reported that distributed from layer II to the white matter, the CCK-PV- the percentage of PV-ir cells among all cortical neurons is co-expressing cells were only found in layer IV. In relatively high (* 14%), only slightly less than that of addition, CCK-ir neurons were not labeled by the antibody GABAergic neurons (* 17%) in human epileptic cortex to SST (Fig. 6B). We performed this double-staining in [54]. In non-epileptic human prefrontal cortex, PV-ir cells tissues from patients #2–4. constitute * 20% of all GABAergic neurons [26]. If GABAergic neurons comprise 20%–30% of the total

123 Q. Zhu et al.: Interneuron Distribution in Human Cortex

Fig. 3 Laminar distribution of TH-ir, NPY-ir, and CCK-ir neurons. neurons in the left anterior temporal cortex of patient #4 (16 years A Laminar distribution of TH-ir neuron across cortical layers in the old). F–H Group data showing the laminar distribution of TH-ir, left frontal lobe of patient #3 (14 years old). Note that TH-ir neurons NPY-ir, and CCK-ir neurons (black) and their density in each layer were located in the deep layers. B Higher magnification of the boxed (blue). I Percentage of cells labeled by different markers in the total area in A. Note the dense distribution of TH-ir fibers in layers I and II. cortical neuron population (black, patients #1, 3, 4, 7, 8, and 9) and C Higher magnification of the boxed area in A. D Example of staining their density in gray matter (blue, patients #7–9). Error bars denote of NPY-ir neurons (patient #3). E Example of staining of CCK-ir SEM.

123 Neurosci. Bull.

Fig. 4 Co-expression of NPY with SST or PV. A–C Double-labeling highly co-localized in layer VI and white matter cells. D Double- for SST and NPY in layers IV (A), VI (B), and the white matter (C). staining for NPY and PV showing that all the three NPY cells were Arrows indicate cells co-expressing NPY and SST; arrowheads PV-negative. Images were from tissues of patient #1 (1.3 years old). indicate cells without co-expression. Note that NPY and SST were

123 Q. Zhu et al.: Interneuron Distribution in Human Cortex

Table 6 Percentages (mean ± SEM) of cells showing co-expression in the total population of PV, SST, TH, NPY, or CCK-positive cells Neurons containing Percentage showing co-localization PV SST TH NPY CCK

PV – – 1.11% ± 0.33% 0 15.6% ± 3.1% SST – – 9.30% ± 1.18% 38.2% ± 3.1% 0 TH 14.6% ± 5.6% 34.2% ± 7.3% – 0 1.79% ± 0.67% NPY 0 97.7% ± 1.5% 0 – 0 CCK 28.0% ± 5.4% 0 0.25% ± 0.09% 0 –

Fig. 5 Co-expression of TH with PV and SST. A Triple-labeling for A but from patient #1 (right temporal lobe, 1.3 years old). TH-positive TH, PV, and SST in tissues from the left frontal lobe (patient #3, 14 cells were more likely to co-express PV. Arrows point to a TH-ir years old). TH-positive cells were more likely to co-express SST. neuron co-expressing PV, while arrowheads indicate a TH-ir neuron Arrows indicate cells with TH and SST co-expression; arrowheads with no PV labeling. point to a TH cell without SST or PV expression. B Labeling as in cortical neuron population [2, 3, 55, 56], the estimated peri-tumor tissue in our experiments, and similar to the PV percentage of PV-ir cells in the cortex would be 4%–6%. In density in cortical tissue without abnormal spiking (35 per agreement with this estimate, our results from non-epileptic mm2)[25] and cryptogenic tissue (* 42 per mm2)[57]. human tissue revealed that * 7% of all cortical neurons However, the density was higher than that in the epilep- showed PV-ir (data not shown), similar to that observed in togenic zone of epilepsy patients (* 20 per mm2)[57], epileptic tissue (7%, Fig. 2 and Table 4). In addition, the suggesting that only PV neurons in the epileptogenic zone laminar distribution pattern of PV-ir neurons in epileptic are vulnerable to epileptic conditions. tissue (Fig. 2 and Tables 2, 3) was also similar to that in Distinct from PV-ir cells, however, other cell types non-epileptic tissue [26]. Since we chose to use tissue might be vulnerable to epileptic conditions. The cell sections showing clear stratiﬁcation (from layer I to the densities of SST-ir and NPY-ir neurons in relatively normal white matter) without any structural abnormality, they tissue have been reported as * 40 and 17 per mm2, represented relatively normal tissue surrounding the respectively [25], much higher than in our experiments (16 epileptogenic zone. Indeed, the PV-ir cell density (* 34 and 2 per mm2, Table 3), suggesting a substantial reduction per mm2, Table 3) was comparable to that in non-epileptic of SST-ir and NPY-ir cells in epileptic conditions. Previous

123 Neurosci. Bull.

Fig. 6 Co-expression of CCK with PV or SST. A Double-labeling for B Double-labeling for SST and CCK in the left anterior temporal CCK and PV in the left anterior temporal cortex of patient #4 (16 cortex of patient #2 (1.6 years old). Note the absence of co-expressing years old). Arrowheads indicate a cell positive for both CCK and PV. cells. Note the absence of PV-ir in the other two CCK-positive cells. studies have shown that NPY neurons are widely dis- not express SST and TH. Previous studies in rodents have tributed throughout the cortex but more frequently in layers shown that only a subpopulation of NPY neurons express II, III, and IV [51, 58, 59]. However, our results revealed SST [49–51]; our results indicated, however, that most that NPY neurons were mainly present in deep layers (IV, NPY cells were positive for SST but negative for PV. We V, and VI) and the white matter. A similar deep-layer also found that subpopulations of TH cells also expressed distribution pattern of TH-ir cells was observed in our PV or SST. In these experiments, we used tissues from immunostaining experiments, consistent with previous different brain regions including the frontal or temporal ﬁndings in human tissue [37, 39, 60]. CCK-ir cells may lobes. Apart from the infant patient (#1) who showed also be subject to alterations in epileptic conditions. The distinct co-expression patterns between TH and PV or SST, CCK mRNA expression level in the rodent cortex increases we found no evident difference in the co-expression of after multiple consecutive kindled seizures [61]. In our different markers, indicating a similar co-expression pat- experiments, we used tissues from infant, teenage, and tern across different brain regions. adult patients (#1, 3, 4, 7, 8, and 9; Table 1) and found, in Together, our results revealed the distribution pattern general, no clear difference in cell proportions and laminar and the overlap of biochemically-identiﬁed cortical distribution patterns for most of the markers, suggesting GABAergic neurons. The results also suggested that the that cell distribution does not change during early devel- density and laminar distribution of PV-ir cells are largely opment. The proportion of TH-positive cells in the whole preserved in epileptic tissue. However, whether other population of cortical neurons in the infant patient was interneuron types (e.g. SST, CCK, NPY, and TH) change relatively lower, possibly due to the high neuronal density in response to epileptic seizures remains to be further in early developmental stages. Moreover, we found that the examined. In addition to the markers used in the current percentages of positive cells in all cortical neurons showed study, non-pyramidal GABAergic neurons in the cortex little difference between patients with different durations of also express other neuropeptides, calcium-binding proteins, epilepsy (Table 1). and neurotransmitter receptor markers, such as vasoactive In the co-expression experiments, we found that NPY-ir intestinal polypeptide, nitric oxide synthase, calretinin, cells did not express PV, TH, or CCK, and CCK cells did calbindin, and 5-hydroxytryptamine 3a receptor. The

123 Q. Zhu et al.: Interneuron Distribution in Human Cortex

Fig. 7 No overlap of TH-, NPY-, and CCK-positive cells. A Double- anterior temporal lobe of patient #6. C As in B, but with double- labeling for NPY and TH in the left anterior temporal lobe of patient staining for TH and CCK (patient #6, 28 years old). Note the absence #2 (1.6 years old). B Double-labeling for NPY and CCK in the left of cells co-expressing any two of the three markers. distribution patterns of these neurons in the human cortex References need to be examined in future studies. 1. Mody I, Pearce RA. Diversity of inhibitory neurotransmission Acknowledgements This work was supported by the National through GABA(A) receptors. Trends Neurosci 2004, 27: Natural Science Foundation of China (31430038 and 81571275). 569–575. 2. Markram H, Toledo-Rodriguez M, Wang Y, Gupta A, Silberberg Open Access This article is distributed under the terms of the G, Wu C. Interneurons of the neocortical inhibitory system. Nat Creative Commons Attribution 4.0 International License (http:// Rev Neurosci 2004, 5: 793–807. creativecommons.org/licenses/by/4.0/), which permits unrestricted 3. Pfeffer CK, Xue M, He M, Huang ZJ, Scanziani M. Inhibition of use, distribution, and reproduction in any medium, provided you give inhibition in visual cortex: the logic of connections between appropriate credit to the original author(s) and the source, provide a molecularly distinct interneurons. Nat Neurosci 2013, 16: link to the Creative Commons license, and indicate if changes were 1068–1076. made. 4. Gonchar Y, Burkhalter A. Three distinct families of GABAergic neurons in rat visual cortex. Cereb Cortex 1997, 7: 347–358. 5. Rudy B, Fishell G, Lee S, Hjerling-Lefﬂer J. Three groups of interneurons account for nearly 100% of neocortical GABAergic neurons. Dev Neurobiol 2011, 71: 45–61.

123 Neurosci. Bull.

6. Wamsley B, Fishell G. Genetic and activity-dependent mecha- 27. Lewis DA, Campbell MJ, Morrison JH. An immunohistochem- nisms underlying interneuron diversity. Nat Rev Neurosci 2017, ical characterization of somatostatin-28 and somatostatin-281-12 18: 299–309. in monkey prefrontal cortex. J Comp Neurol 1986, 248: 1–18. 7. Varga C, Tamas G, Barzo P, Olah S, Somogyi P. Molecular and 28. Seney ML, Tripp A, McCune S, Lewis DA, Sibille E. Laminar electrophysiological characterization of GABAergic interneurons and cellular analyses of reduced somatostatin gene expression in expressing the transcription factor COUP-TFII in the adult human the subgenual anterior cingulate cortex in major depression. temporal cortex. Cereb Cortex 2015, 25: 4430–4449. Neurobiol Dis 2015, 73: 213–219. 8. Kepecs A, Fishell G. Interneuron cell types are fit to function. 29. Chen G, Zhang Y, Li X, Zhao X, Ye Q, Lin Y, et al. Distinct Nature 2014, 505: 318–326. inhibitory circuits orchestrate cortical beta and gamma band 9. Gonchar Y, Wang Q, Burkhalter A. Multiple distinct subtypes of oscillations. Neuron 2017, 96: 1403–1418. GABAergic neurons in mouse visual cortex identified by triple 30. Takada N, Pi HJ, Sousa VH, Waters J, Fishell G, Kepecs A, et al. immunostaining. Front Neuroanat 2007, 1: 3. A developmental cell-type switch in cortical interneurons leads to 10. Silberberg G, Markram H. Disynaptic inhibition between neo- a selective defect in cortical oscillations. Nat Commun 2014, 5: cortical pyramidal cells mediated by Martinotti cells. Neuron 5333. 2007, 53: 735–746. 31. Bartos M, Vida I, Jonas P. Synaptic mechanisms of synchronized 11. Silberberg G. Polysynaptic subcircuits in the neocortex: spatial gamma oscillations in inhibitory interneuron networks. Nat Rev and temporal diversity. Curr Opin Neurobiol 2008, 18: 332–337. Neurosci 2007, 8: 45–56. 12. Ren M, Yoshimura Y, Takada N, Horibe S, Komatsu Y. 32. Yekhlef L, Breschi GL, Lagostena L, Russo G, Taverna S. Specialized inhibitory synaptic actions between nearby neocor- Selective activation of parvalbumin- or somatostatin-expressing tical pyramidal neurons. Science 2007, 316: 758–761. interneurons triggers epileptic seizurelike activity in mouse 13. Li T, Tian C, Scalmani P, Frassoni C, Mantegazza M, Wang Y, medial entorhinal cortex. J Neurophysiol 2015, 113: 1616–1630. et al. Action potential initiation in neocortical inhibitory 33. Cheung A, Newland PL, Zaben M, Attard GS, Gray WP. interneurons. PLoS Biol 2014, 12: e1001944. Intracellular nitric oxide mediates neuroproliferative effect of 14. Marin O. Interneuron dysfunction in psychiatric disorders. Nat neuropeptide y on postnatal hippocampal precursor cells. J Biol Rev Neurosci 2012, 13: 107–120. Chem 2012, 287: 20187–20196. 15. Nakazawa K, Zsiros V, Jiang Z, Nakao K, Kolata S, Zhang S, 34. Benarroch EE. Neuropeptide Y: its multiple effects in the CNS et al. GABAergic interneuron origin of schizophrenia pathophys- and potential clinical significance. Neurology 2009, 72: iology. Neuropharmacology 2012, 62: 1574–1583. 1016–1020. 16. Lewis DA. GABAergic local circuit neurons and prefrontal 35. Colmers WF, El Bahh B. Neuropeptide Y and epilepsy. Epilepsy cortical dysfunction in schizophrenia. Brain Res Brain Res Rev Curr 2003, 3: 53–58. 2000, 31: 270–276. 36. Raghanti MA, Conley T, Sudduth J, Erwin JM, Stimpson CD, 17. Rubenstein JL, Merzenich MM. Model of autism: increased ratio Hof PR, et al. Neuropeptide Y-immunoreactive neurons in the of excitation/inhibition in key neural systems. Genes Brain Behav cerebral cortex of humans and other haplorrhine primates. Am J 2003, 2: 255–267. Primatol 2013, 75: 415–424. 18. van Rijckevorsel K. Cognitive problems related to epilepsy 37. Gaspar P, Berger B, Febvret A, Vigny A, Krieger-Poulet M, syndromes, especially malignant epilepsies. Seizure 2006, 15: Borri-Voltattorni C. Tyrosine hydroxylase-immunoreactive neu- 227–234. rons in the human cerebral cortex: a novel catecholaminergic 19. Alherz F, Alherz M, Almusawi H. NMDAR hypofunction and group? Neurosci Lett 1987, 80: 257–262. somatostatin-expressing GABAergic interneurons and receptors: 38. Ikemoto K, Kitahama K, Nishimura A, Jouvet A, Nishi K, Arai R, A newly identified correlation and its effects in schizophrenia. et al. Tyrosine hydroxylase and aromatic L-amino acid decar- Schizophr Res Cogn 2017, 8: 1–6. boxylase do not coexist in neurons in the human anterior 20. Scharfman HE. The neurobiology of epilepsy. Curr Neurol cingulate cortex. Neurosci Lett 1999, 269: 37–40. Neurosci Rep 2007, 7: 348–354. 39. Benavides-Piccione R, DeFelipe J. Distribution of neurons 21. Liu YQ, Yu F, Liu WH, He XH, Peng BW. Dysfunction of expressing tyrosine hydroxylase in the human cerebral cortex. hippocampal interneurons in epilepsy. Neurosci Bull 2014, 30: J Anat 2007, 211: 212–222. 985–998. 40. Wyeth MS, Zhang N, Mody I, Houser CR. Selective reduction of 22. Marco P, Sola RG, Pulido P, Alijarde MT, Sanchez A, Ramon y cholecystokinin-positive basket cell innervation in a model of Cajal S, et al. Inhibitory neurons in the human epileptogenic temporal lobe epilepsy. J Neurosci 2010, 30: 8993–9006. temporal neocortex. An immunocytochemical study. Brain 1996, 41. Morozov YM, Freund TF. Postnatal development and migration 119 (Pt 4): 1327–1347. of cholecystokinin-immunoreactive interneurons in rat hippocam- 23. Martinello K, Sciaccaluga M, Morace R, Mascia A, Arcella A, pus. Neuroscience 2003, 120: 923–939. Esposito V, et al. Loss of constitutive functional gamma- 42. Pawelzik H, Hughes DI, Thomson AM. Physiological and aminobutyric acid type A-B receptor crosstalk in layer 5 morphological diversity of immunocytochemically defined par- pyramidal neurons of human epileptic temporal cortex. Epilepsia valbumin- and cholecystokinin-positive interneurones in CA1 of 2018, 59: 449–459. the adult rat hippocampus. J Comp Neurol 2002, 443: 346–367. 24. Wonders CP, Anderson SA. The origin and specification of 43. Sun C, Sun J, Erisir A, Kapur J. Loss of cholecystokinin- cortical interneurons. Nat Rev Neurosci 2006, 7: 687–696. containing terminals in temporal lobe epilepsy. Neurobiol Dis 25. Gonzalez-Albo MC, Elston GN, DeFelipe J. The human temporal 2014, 62: 44–55. cortex: characterization of neurons expressing nitric oxide 44. Najm IM, Sarnat HB, Blumcke I. Review: The international synthase, neuropeptides and calcium-binding proteins, and their consensus classification of Focal Cortical Dysplasia - a critical glutamate receptor subunit profiles. Cereb Cortex 2001, 11: update 2018. Neuropathol Appl Neurobiol 2018, 44: 18–31. 1170–1181. 45. Mullen RJ, Buck CR, Smith AM. NeuN, a neuronal specific 26. Fish KN, Rocco BR, Lewis DA. Laminar distribution of subsets nuclear protein in vertebrates. Development 1992, 116: 201–211. of GABAergic axon terminals in human prefrontal cortex. Front 46. Wang G, He F, Xu Y, Zhang Y, Wang X, Zhou C, et al. Neuroanat 2018, 12: 9. Immunopotentiator thymosin alpha-1 promotes neurogenesis and

123 Q. Zhu et al.: Interneuron Distribution in Human Cortex

cognition in the developing mouse via a systemic Th1 bias. 54. Thom M, Martinian L, Parnavelas JG, Sisodiya SM. Distribution Neurosci Bull 2017, 33: 675–684. of cortical interneurons in grey matter heterotopia in patients with 47. Tansey KE, Hill MJ. Enrichment of schizophrenia heritability in epilepsy. Epilepsia 2004, 45: 916–923. both neuronal and glia cell regulatory elements. Transl Psychiatry 55. Beaulieu C. Numerical data on neocortical neurons in adult rat, 2018, 8: 7. with special reference to the GABA population. Brain Res 1993, 48. DeFelipe J. Anatomical and molecular heterogeneity of cortical 609: 284–292. gabaergic interneurons. In: Grillner S, Graybiel AM (Eds.). 56. DeFelipe J. Chandelier cells and epilepsy. Brain 1999, 122 (Pt Microcircuits: The Interface Between Neurons And Global Brain 10): 1807–1822. Function. Cambridge, MA: The MIT Press, 2006. 57. Medici V, Rossini L, Deleo F, Tringali G, Tassi L, Cardinale F, 49. Karagiannis A, Gallopin T, David C, Battaglia D, Geoffroy H, et al. Different parvalbumin and GABA expression in human Rossier J, et al. Classiﬁcation of NPY-expressing neocortical epileptogenic focal cortical dysplasia. Epilepsia 2016, 57: interneurons. J Neurosci 2009, 29: 3642–3659. 1109–1119. 50. Kawaguchi Y, Kubota Y. GABAergic cell subtypes and their 58. Kubota Y, Hattori R, Yui Y. Three distinct subpopulations of synaptic connections in rat frontal cortex. Cereb Cortex 1997, 7: GABAergic neurons in rat frontal agranular cortex. Brain Res 476–486. 1994, 649: 159–173. 51. Hendry SH, Jones EG, Emson PC. Morphology, distribution, and 59. Kuljis RO, Rakic P. Distribution of neuropeptide Y-containing synaptic relations of somatostatin- and neuropeptide Y-im- perikarya and axons in various neocortical areas in the macaque munoreactive neurons in rat and monkey neocortex. J Neurosci monkey. J Comp Neurol 1989, 280: 383–392. 1984, 4: 2497–2517. 60. Benavides-Piccione R, DeFelipe J. Different populations of 52. Sherwood CC, Raghanti MA, Stimpson CD, Spocter MA, Uddin tyrosine-hydroxylase-immunoreactive neurons deﬁned by differ- M, Boddy AM, et al. Inhibitory interneurons of the human ential expression of nitric oxide synthase in the human temporal prefrontal cortex display conserved evolution of the phenotype cortex. Cereb Cortex 2003, 13: 297–307. and related genes. Proc Biol Sci 2010, 277: 1011–1020. 61. Zhang LX, Smith MA, Kim SY, Rosen JB, Weiss SR, Post RM. 53. Wittner L, Eross L, Czirjak S, Halasz P, Freund TF, Magloczky Changes in cholecystokinin mRNA expression after amygdala Z. Surviving CA1 pyramidal cells receive intact perisomatic kindled seizures: an in situ hybridization study. Brain Res Mol inhibitory input in the human epileptic hippocampus. Brain 2005, Brain Res 1996, 35: 278–284. 128: 138–152.

123